Engine Noise Diagnosis Combining Kurtosis and Modulation Frequency Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current engine combustion abnormality diagnosis methods, such as Kurtosis analysis, fail to accurately determine mechanical abnormalities caused by rotational excitation, as they only consider energy distribution and not the influence of rotation and combustion excitations.
Innovation Solution
A mechanical diagnosis method that combines Kurtosis analysis with Modulation Frequency analysis to classify noise patterns in high-frequency regions, using a diagnosis controller to calculate Kurtosis and Modulation Frequency Indices from engine noise data, and determine abnormality states by applying predetermined thresholds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If Kurtosis analysis is used for combustion abnormality diagnosis, then the energy distribution analysis is simple and computationally efficient, but the analysis cannot detect mechanical abnormalities caused by rotational excitation
Solution Approach 1:
The patent combines Kurtosis analysis with Modulation Frequency analysis to create a comprehensive diagnostic system. The Kurtosis analysis component provides simple energy distribution analysis, while the Modulation Frequency analysis component detects rotational excitation patterns. By merging these two analysis methods, the system achieves both computational efficiency and accurate detection of mechanical abnormalities that would be missed by Kurtosis analysis alone.
2Ease of operation
If only energy distribution analysis is performed, then the diagnostic process is straightforward, but the influence of rotation and combustion excitations cannot be distinguished
Solution Approach 1:
The patent segments the noise signal analysis into distinct frequency components using Modulation Frequency analysis. By dividing the frequency spectrum and analyzing specific frequency bands separately, the system can identify and distinguish between rotational excitation patterns and combustion excitation patterns. This segmentation preserves excitation influence information that would be lost in overall energy distribution analysis.
3Device complexity
If the diagnostic system uses only Kurtosis index, then the threshold application is simple, but problem samples with rotation excitation influence cannot be accurately classified
Solution Approach 1:
The patent creates a multi-functional diagnostic system where the Modulation Frequency analysis component serves multiple purposes: it detects rotational excitation, identifies mechanical abnormalities, and provides additional diagnostic information that complements the Kurtosis analysis. This universal approach enhances the reliability of combustion abnormality determination by ensuring that problem samples with rotation excitation influence are accurately classified, while maintaining relatively simple system structure through the use of established analysis methods.
Data Source
AI summary
A mechanical diagnosis method for combustion abnormality using engine noise includes: calculating an Energy K and a Loudness standard deviation index (NSTD) with Kurtosis analysis by a diagnosis controller 30 from noise data measured together with a signal component by rotation excitation of an engine 10; calculating a plurality of order frequency peak order component values by a Modulation Frequency Transform; and distinguishing a cylinder where abnormal combustion occurs from a cylinder where normal combustion occurs by applying a predetermined threshold to these calculated values, thereby classifying, by Modulation Frequency analysis, problem samples of the rotation excitation and combustion excitation influence of the engine in which the abnormality state determination of the engine has been difficult only with energy distribution while overcoming the limitation of Kurtosis analysis.